You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I end up with a total charge of 2e-6 for a lipid, which is not bad. If then I reuse the generated topology as input for GROMACS simulation of a membrane with, say 1000 lipids, the total membrane charge is around 0.002, which is maybe significant (at least it causes a GROMACS error).
The origin of a problem is two-fold, as I see it:
The charges are not normalize when Electrostatics collection is created from OpenMM
Rounding error is not addressed, when writing GROMACS topology
I think that working on both aspects can be worth considering. Meanwhile I suggest a simple solution (#1167 ) for ensuring that rounding errors upon writing GROMACS topologies are not present.
The text was updated successfully, but these errors were encountered:
Description
I am using interchange to write topologies for lipids generated with openmm
Modeller.addMembrane
method.Overall, my code looks like that:
As input I am using the following membrane protein structure (let me know if access is restricted).
I end up with a total charge of
2e-6
for a lipid, which is not bad. If then I reuse the generated topology as input for GROMACS simulation of a membrane with, say 1000 lipids, the total membrane charge is around 0.002, which is maybe significant (at least it causes a GROMACS error).The origin of a problem is two-fold, as I see it:
Electrostatics
collection is created fromOpenMM
I think that working on both aspects can be worth considering. Meanwhile I suggest a simple solution (#1167 ) for ensuring that rounding errors upon writing GROMACS topologies are not present.
The text was updated successfully, but these errors were encountered: